A model for optimal reaction norms: the case of the pregnant garter snake and her temperature-sensitive embryos.

We present a model to test Osgood's ( 1978 ) proposition that viviparous snakes have optimal reaction norms for temperature-sensitive meristic traits, such as scale counts. Our model predicts that traits that are subject to temperature effects during development will evolve a flat or [Formula: see text]-shaped reaction norm (average scale count as a function of developmental temperature). We tested this prediction by maintaining 67 female garter snakes (Thamnophis elegans) at eight different constant temperatures (21 degrees -33 degrees C) during pregnancy and making a series of scale counts on their newborn offspring (n = 491). To insure that the experimental temperatures were ecologically relevant, we used automated radiotelemetry to record the body temperature of pregnant, free-ranging females. The resulting temperature data allowed us to test the prediction that the inflection points of reaction norms would correspond to the average temperature experienced by embryos in the field. In line with predictions of the Osgood model, reaction norms were flat or U-shaped. In the case of U-shaped reaction norms, the inflection point of the curves corresponded to the average temperature imposed on embryos by free-ranging females. In contrast to some past studies, none of the standard scale scores (ones commonly used in systematics) showed significant temperature effects in either sex. Reaction norms were flat. In contrast, incidences of various abnormalities showed U-shaped reaction norms. Temperature effects were more pronounced in males than in females. The results have implications for systematics and for the evolution of canalization and phenotypic plasticity.